Switching Delay Analysis for Two Neuronal Toggle Switch Designs: Direct and Staged Mutual Inhibition

Direct Mutual Inhibition is known to be the primary mechanism behind bi-stabile behavior in biological neuronal networks to implement a neuronal toggle switch, similar to flip-flops in electronics. With a view to enabling the implementation of this pervasive behavior in neuromorphic computing, we have previously used computational models to conclude that direct mutual inhibition suffers from a common mode failure, and have proposed an alternative design, staged mutual inhibition, that ensures correct switching. In this work, we provide further analysis for these two designs with respect to switching delay as a performance metric, similar to hold-time considerations in electronics. Our timing analysis shows that staged mutual inhibition incurs slightly larger delay compared with direct mutual inhibition, which is attributed to its more sophisticated design. We perform simulations to validate our quantitative delay analysis.